Henry Throop!!University of Pretoria !

&!Planetary Science Institute!

Tucson, Arizona!!

Collaborators:!!

John Bally (U. Colorado)!Nickolas Moeckel (Cambridge)!

!!

University of the Witwatersrand, Johannesburg April 8, 2014

Planet Formation in Dense Star Clusters

3

Orion Constellation (visible light)

Orion constellation!H-alpha

Orion constellation!H-alpha

Orion Molecular Clouds!>105 Msol 100 pc long

Orion core (visible light)

Orion core (visible light)

Orion Star Forming Region!!• Closest bright star-forming region to Earth!• Distance ~ 1500 ly!• Age ~ 10 Myr!• Radius ~ few ly!• Mean separation ~ 104 AU

Orion Trapezium cluster

O/B stars

Low mass stars; Disks with tails

• Largest Orion disk: 114-426, D ~ 1200 AU !• Dust grains in disk are grey, and do not redden light as they extinct it!• Dust grains have grown to a few microns or greater in < 1 Myr!

Largest Orion disk: 114-426, diameter 1200 AU !

Star Formation

17

1961 view:!“Whether we've ever seen a star form or not is still debated. The next slide is the one piece of evidence that suggests that we have. Here's a picture taken in 1947 of a region of gas, with some stars in it. And here's, only two years later, we see two new bright spots. The idea is that what happened is that gravity has...”!

!Richard Feynman, Lectures on Physics

2000s view: !Infrared detectors have allowed us to directly see thousands of star forming -- nearly everywhere that we see an IR source. 1000+ young stars in Orion alone.!!

Whether we’ve ever seen a planet form or not is the current question!

Star Cluster Formation Star Formation Planet Formation

Circumstellar Disks In Orion

• 100+ disks directly observed, diameters 100-1200 AU!• 80%+ of stars in Orion show evidence for having disks!! These stars are too distant and young to directly search for planets… but we want to study the environment and processes to understand the planets which would be produced in these dense clusters -- and therefore throughout the galaxy.

Regions of Star FormationLarge Dense Clusters:! Orion

Small Sparse Clusters: ! Taurus

# of stars 103

104

(Orion)

10 -

OB stars Yes No

Distance 450 pc (Orion) 140 pc (Taurus)

Fraction of stars that form here

70-90% 10-30%

Distance between stars

5000 AU 20,000 AU

Dispersal lifetime Few Myr

% of stars with disks > 80%

Orion: Hot, Dense, Massive!!Most stars form in large clusters.

Taurus:!Dark, Small, Cold!!Most planet formation models study small clusters.

Where did our Sun form?

• We don’t know! The Sun is an isolated star today.!• 90% of stars formed in clusters!• But just 1% remain in clusters now.!• Stellar motions can be back-integrated for 100 Myr, but not 10 Gyr.!

• 60Fe isotopes suggest Sun was born in a large cluster, few pc away from a supernova

Planet Formation - Classical Model

21

!Cloud core collapses due to self-gravity! 10,000 AU, 1 Msol!!!!Disk flattens; grains settle to midplane!Planet cores grow! Disk Mass: ‘Minimum Mass Solar Nebula’! MMSN = 0.01 Msol!

Star Mass: ~ 1 Msol!!Terrestrial planets form!Jovian planets accrete gas!!!Disk disperses!Solar System complete after ~ 5-10 Myr!!!

W. Hartmann

Work we have done involves ...!!!

– UV photo-evaporation from massive stars!!!

– Interaction with cluster gas!!!

– Close stellar encounters!!!

– Organics and UV photolysis from massive stars

How does Cluster Environment affect Disk Evolution?

Throop 2000; Bally et al 2005; Throop & Bally 2005; Throop & Bally 2008; Moeckel & Throop 2009; Throop & Bally 2010; Pichardo et al 2010; Throop 2011.

Bondi-Hoyle Accretion

• Cool molecular H2 from cluster ISM accretes onto disks!• Accretion flow is onto disk, not star.!• Accretion is robust against stellar winds, radiation pressure, turbulence.!• This accretion is not considered by existing Solar System formation

models!! 1 MMSN = 1 ‘Mimimum Mass Solar Nebula’ = 0.01 MSol

26Accretion radius ~ 1000 AU Accretion rate ~ 10-8 M! yr-1Accretion rate ~ 1 MMSN / Myr

2 RB

Timescale of Star Formation

27

Gas Accretion + N-Body Cluster Simulations

NBODY6 code (Aarseth 2003)!

!Stars:!• N=1000!• Mstar = 500 M!!

• Kroupa IMF!• R0 = 0.5 pc!!Gas:!• Mgas = 500 M!!

• R0 = 0.5 pc!• Disperses with

timescale 2 Myr

Throop & Bally 2008

29

Following trajectory of one star of 3000 from N-body simulation...

BH Accretion: History of individual star

• Star+disk accretes 5% of own mass in 5 Myr.!• Accretion is episodic!

– Highest at core: High velocity but high density

BH Accretion: History of individual star

Results of N-Body sims

!!

• Typical mass accreted by disks surrounding Solar-mass stars is 1 MMSN per Myr!

• Accretion occurs for several Myr, until cluster disperses or cloud is ionized!

Observations of accretion in young stars

• Accretion is seen onto hundreds young stars in molecular clouds.!!

• Varies with stellar mass: dM/dt ~ M2 !

!• Accretion is ~ 0.01 M! Myr-1 for 1M!!!

• Source of the accretion is unknown!

Observations of accretion in young stars

Text

We propose: accretion onto young stars may be due to ISM accretion onto their disks

• Accretion is seen onto hundreds young stars in molecular clouds.!!

• Varies with stellar mass: dM/dt ~ M2 !

!• Accretion is ~ 0.01 M! Myr-1 for 1M!!!

• Source of the accretion is unknown!

Throop & Bally 2008

Consequences of Tail-End Accretion

• Disks may accrete many times their own mass in a few Myr. !• Disks may still be accreting gas at >5 Myr, after

planetesimals form, and maybe after giant planet cores form.!• Disk may be ‘rejuvenated’ after being partially lost!• Final composition of disk may be different than star!

– There may be no ‘Solar Nebula Composition’!– Isotopes may not be diagnostic of solar vs. extrasolar material

34

Throop & Bally 2008, AJ!

Accretion of ‘polluted’ ISM

• Stars of same age/position/type in Orion show metallicities that vary by up to 10x in Fe, O, Si, C!

• Could stars have accreted metallic ‘veneers’ by passing through nearby molecular clouds, contaminated with supernova ejecta?!

• 20 MSol SN produces 4 MSol O

35

Late accretion may cause the composition of a stars and their disks to be different! There may be no ‘Solar Nebula Composition.’ Even in our Solar System, there is a lot of variation : isotope ratios.

Cunha et al 2000

36

Mol

ecul

ar c

loud

s

37

Mol

ecul

ar c

loud

s

38

Mol

ecul

ar c

loud

s

Ionized HII region

39

Mol

ecul

ar c

loud

s

Ionized HII region

40

Mol

ecul

ar c

loud

s

Ionized HII region

41

Mol

ecul

ar c

loud

s

Ionized HII region

42

Mol

ecul

ar c

loud

s

Ionized HII region

43

Mol

ecul

ar c

loud

s

Ionized HII region

44

Mol

ecul

ar c

loud

s

Ionized HII region

45

Mol

ecul

ar c

loud

s

Ionized HII region

46

Mol

ecul

ar c

loud

s

Ionized HII region

47

Mol

ecul

ar c

loud

s

Ionized HII region

48

Mol

ecul

ar c

loud

s

Ionized HII region

49

Mol

ecul

ar c

loud

s

Orion constellation!H-alpha

Orion Molecular Clouds!>105 Msol 100 pc long

Jupiter vs. the Sun

If the Sun and Jupiter both formed from the same cloud, why are they made of such different stuff?

Jupiter’s Atmosphere

• Mass Spectrometer aboard Galileo Probe!• Measured to ~20 bars!• Found Jupiter atmosphere to be 2-6x

higher in metals vs. Sun, when normalized to H.!

– C, S, Ar, Kr, Xe!– All these are stable and long-lived: enrichment

was a complete surprise!!– vesc = 45 km/sec!

• GPMS likely passed through ‘dry spot’ on Jupiter.!

• Several explanations proposed:!– Noble gases may be enhanced by

freeze-out onto ices. But requires extremely cold disk < 30K (Guillot, Hersant, Lunine).!

– Jupiter may be H-depleted, and S could be a better reference (Lodders 2004). 53

Jupiter ‘Polluted Accretion’ model

We propose a crazy idea for Jupiter’s composition:!!!1. Solar System forms in a large star cluster.!!2. Massive stars pollute ISM with heavy elements.!

SNs and massive stellar winds convert H into C, N, S, etc.!

!3. ‘Pollution’ from massive stars is accreted onto Jupiter.!

Accretion from ISM -> Solar Nebula Disk -> Jupiter!Sun’s metallicity is not affected, only Jupiter’s

54Throop & Bally 2010 (Icarus)

Observed Jupiter Composition

Can Jupiter’s measured enhancement be explained by accretion of heavy elements from the ISM?

55

Jupiter ‘Polluted Accretion’ model

• Data: Galileo Probe!• Model: Accretion from ISM!

– 87% Solar nebula material!– 9% Stellar winds from 20 M! star (provides C, N)!– 4% SN from 25 M! star (provides S, Ar, Kr, Xe)!– Requires total of ~0.13 MJ of accretion to explain Jupiter’s

current metallicity.!– Bondi-Hoyle accretion supplies 10 MJ of accretion per Myr --

plenty of mass, and with the right chemistry!

• Evidence for a heterogeneous nebula is not new!!!!!!!!!!!!

• Heterogeneity between Jupiter and Sun is a natural extension to that already observed in meteorites (but much bigger).

Jupiter ‘Polluted Accretion’ model

Dauphas et al 2002:!

“'Mb isotope abundances were heterogeneously distributed in the Solar System’s parental molecular cloud, and the large-scale variations we observed were inherited from the interstellar environment where the Sun was born.”

Ranen & Jacobsen 2006:!

“There are resolvable differences between the Earth and carbonaceous chondrites that are most likely caused by incomplete mixing of r- and s-process nucleosynthetic components in the early Solar System.”

Trinquier et al 2007:!

“Preservation of the 54Cr heterogeneity in space and time (several Myr) motivates us to speculate that late stellar input(s) could have been significant contributions to inner nebular Cr reservoirs...”

Throop & Bally 2010 (Icarus, in review)

SPH Sims: BH Acc onto 100 AU disk

58

10,000 years!0.01 solar masses!v ~ 1 km/sec!!!Moeckel & Throop 2009 (AJ)

Close Stellar Encounters

• Typical distances today ~ 10,000 AU!!

• C/A strips disks to 1/3 the closest-approach distances (Hall et al 1996)!!

• Question: What is the minimum C/A distance a disk encounters as it moves through the cluster for several Myr?

HST 16 200 AU diameter

! 0.3 ly to O star

Close Approach History - Typical 1 M! Star

• Star has 5 close approaches at < 2000 AU.!• Closest encounter is 300 AU at 8 Myr!

• Too late to do any damage

Throop & Bally 2008; also Adams et al 2006

Photo-Evaporation in Orion

• Disks surrounding solar-type stars are heated by UV-bright stars.!!

• Gas is heated and removed from disk on 1-10 Myr timescales.!!

• If disk is removed quickly, we can’t form planets!!!

Triggered Planet Formation?

Photo-evaporation removes gas and allows gravitational instability to form planetesimals.

!!!!!!! Throop & Bally 2005

Effects of Photo-Evaporation on Planet Formation

Solar System-like disks are removed in 1-10 Myr. Effects on...!!

• Kuiper Belt (> 40 AU): UV removes volatiles and small grains. Kuiper belts and Oort clouds may be rare! Or, they may be formed easily and quickly thru triggering.!!

• Giant Planets (5-40 AU): Gas is rapidly removed from disk: If you want to build Jupiters in Orion, do it quickly! (e.g., Boss models).!!

• Terrestrial Planets (1-5 AU): Safe against photo-evaporation since it’s hard to remove gas from 1 AU.

Flux Received onto a Disk vs. Time

Punctuated equilibrium

• Flux received by disk varies by 1000x as it moves through the cluster : Freeze-Broil-Freeze-Broil!

• Peak flux approaches 107 G0.!• Most of the flux is deposited during brief but intense

close encounters with core.!• There is no ‘typical UV flux.’!• Photo-evap models assume steady UV flux. But if UV is

not steady, then other processes (viscous, grain growth) can dominate at different times and dramatically change the disk. Throop, in prep

Planet Formation - Classical Model

67

!Cloud core collapses due to self-gravity! 10,000 AU, 1 M!! !!!!Disk flattens; grains settle to midplane!Planet cores grow!!!Terrestrial planets form!Jovian planets accrete gas!!!Disk disperses!Solar System complete after ~ 5-10 Myr!!! W. Hartmann

Planet Formation - Classical Model

68

Cloud is heterogeneous and polluted!Cloud core collapses due to self-gravity! 10,000 AU, 1 M!!Cloud inherits composition from nearby SN!!!!Disk flattens; grains settle to midplane!Planet cores grow!Disk is photo-evaporated by UV stars!Disk is injected with 60Fe from nearby SNs!Terrestrial planets form!Jovian planets accrete gas!(Disk is stripped due to close approaches)!Disk accretes gas from environment!Disk disperses and is photo-evaporated!Solar System complete after ~ 5-10 Myr!!! W. Hartmann

modified

Randomness as a factor in Disk Evolution

• Disk outcome depends not just on its ingredients, but on its individual history.!

• If we try to predict what will form around individual stars or disks, we’re doomed to fail!!

• Disk systems are individuals, they interact with their environment, and random events and timing matter:!– How much stuff was photo-evaporated by UV?!– How hot was the disk, and how viscous, and how did its

surface density evolve?!– How strong, when, and how many times did UV hit it?!– What SN events occurred? How did they contaminate the

disk?!– What molecular clouds did disk pass through? What material

was accreted? Onto inner disk, or outer?!– Do planetesimals form before, or after, photo-evaporation

starts?!• There is no ‘typical’ disk, and no ‘typical’ planetary system, even if

starting from the same initial disk structure and ingredients.

70

The End